
% TABLE 1
\newcommand\TabRandCheck[1]{%
  \begin{table}[#1]
    \setlength{\extrarowheight}{4pt}
    \begin{center}
      \hspace*{-0in}
      \setlength{\tabcolsep}{2.75pt}
      \begin{tabular}{%
        r%
        p{.67em}%
        >{{\hspace*{0em}}}N{3}{2}%
        p{.67em}%
        >{{\hspace*{0em}}}N{1}{2}%
        p{.67em}%
        >{{\hspace*{0em}}}N{0}{2}%
      }
        && \multicolumn{1}{c}{$\chi^2$} &&
           \multicolumn{1}{c}{F}        &&
           \multicolumn{1}{c}{p}\tabularnewline
        \cmidrule(lr){3-3}\cmidrule(lr){5-5}\cmidrule(lr){7-7}
                           Name of Chief Justice && 473.61 &&        && .83\tabularnewline
               How many justices currently serve && 361.80 &&        && .16\tabularnewline
                 How many justices usually serve && 360.02 &&        && .18\tabularnewline
                  How many women currently serve && 375.09 &&        && .07\tabularnewline
                  Name of Senate Majority Leader && 342.26 &&        && .40\tabularnewline
                         Term length of justices && 356.40 &&        && .21\tabularnewline
                         How are justices chosen &&        &&    .98 && .53\tabularnewline
        Can one watch lawyers argue at the Court &&        &&    .96 && .58\tabularnewline
                   Who decides constitutionality &&        &&   1.08 && .30\tabularnewline
           What happens when there is a tie vote &&        &&   1.05 && .36\tabularnewline
                          When can the Court act &&        &&    .95 && .60\tabularnewline
                    Can justices ever be removed &&        &&   1.08 && .29\tabularnewline
        \bottomrule
      \end{tabular}
      %\captionsetup{margin={.75in, .75in}}
      \caption{%
        \label{TabRandCheck}%
        \textit{Randomization checks.} The table reports separate randomization checks for each of our 12 questions. For each question, treatment condition was regressed on age, race, gender, education, and state of residence. The first six questions had five treatment conditions, and for these questions, we used multinomial logistic regression. The six remaining questions had only two treatment conditions (``three response options'' and ``five response options''), and in these cases, we used OLS.
      }
    \end{center}
  \end{table}
}%


